Proceedings of the 7th International Conference on Gas Hydrates (ICGH 2011),
Edinburgh, Scotland, United Kingdom, July 17-21, 2011.
INVESTIGATION ON A NOVEL REACTOR FOR GAS HYDRATE
PRODUCTION
Lucia Brinchi∗ , Beatrice Castellani, Franco Cotana, Mirko Filipponi, Federico Rossi
Department of Industrial Engineering, University of Perugia
Via G. Duranti, Perugia 06125
ITALY
Gianfranco Savelli
CEMIN, University of Perugia
Via Elce di Sotto, 8, 06123 Perugia
ITALY
ABSTRACT
Gas hydrates have a large capacity for the storage of gases and are an attractive method for gas
filtration and transportation. However, hydrate formation is usually controlled by the rate of
crystallization and thus it can be a slow process. A technology that will rapidly and continuously
form hydrate is necessary for scale-up.
A novel reactor with an inner volume of 25 l has been designed, built and installed to form
hydrates by injection of water through spray nozzles from the top of the methane-filled reactor.
This method allows to maximize interfacial area between reactants and to minimize mass transfer
barriers and thermal effects that negatively affect conversion of reactants into hydrate. A system
was developed to remove heat released during hydrate formation, and to reach a good level of
temperature control inside the reactor. A novel system for hydrate recovering and unloading was
also designed. The objective of experimentation is to lower energetic costs of hydrate formation,
also through surfactant promotion. Anionic surfactant sodium dodecyl sulphate, SDS, was tested.
Results of a first set of experimental applications of the reactor for rapid methane hydrate
formation are presented.
Keywords: methane hydrate, surfactant, scale-up, water spraying
NOMENCLATURE
n number of moles
P pressure [Pa]
R universal gas constant
SDS sodium dodecyl sulfate
t time [min]
T temperature [K]
V volume [l]
Z compressibility factor
INTRODUCTION.
Gas hydrates are crystalline compounds formed by
the hydrogen-bonded water molecules in lattice
∗
structure that is stabilized by encapsulating a small
guest molecule, such as methane or ethane [1].
Methane hydrates gives unique gas storage
properties, as the hydrates can contain 150-180
V/V (standard temperature and pressure).
Actually, after the discovery of hydrate selfpreservation which allows hydrates to remain
metastable at few degrees below the ice point at
atmospheric pressure [2-3], scientists have become
interested in studying the storage and
transportation of gas in the form of hydrate.
Research in this area started in the early 1990s by
Gudmundsson and his group [3-4], and thereafter
Corresponding author: Phone: +39 075 5051704 Fax 1782206856 E-mail: [email protected]
several other researchers have reported results in
this area [5-7]. The storage and transport of natural
gas in methane hydrates does have some
advantages over some of the conventional methods
for methane storage. For instance two of the nonpipeline methods of methane storage consist in
liquefying the natural gas (LNG) or in
compressing it (CNG). Economic feasibility
studies have been carried out especially by
Japanese industrial researchers, and show how the
cost of NGH is lower than that of LNG of the
same scale, concluding that NGH is advantageous
over LNG for gas transportation of amounts less
that 1 million ton per annum [7]. Moreover, recent
studies based on calculations of the total energy
involved in the three technologies (NGH, LNG
and CNG), from preparation of methane in its
transportable form to its final use, show how LNG
process is less energy demanding but by far more
expensive when compared to storage and transport
of gas hydrates, whereas CNG is comparable to
the use of gas hydrates from energy and
economical points of view, but the technology is
nowadays developed only for small-scale systems
[8]. The storage and transportation of gas as NGH
is appealing for industrial utilization also because
of its high safety.
Nevertheless, the feasibility of using chlathrate
hydrates for industrial purposes has long been an
intriguing subject of engineering studies [9-10].
The main issues are the slow formation rates,
unreacted interstitial water as a large percentage of
the hydrate mass, reliability of hydrate storage
capacity and economy of process scale-up. The
lack of a technology that allows the production of
hydrate in a continuous manner suitable for scaleup to industrial settings has hindered the utilization
of gas hydrates in the above application, and also
in others, such as, biogas refining, gas separation
and capture, residential air conditioning,
desalinization of seawater. Hydrate formation is
always the first process in any of these yet-to-be
established technologies, and it may control the
rate of the serial processes composing the
technology. Thus, how to efficiently form hydrates
is a technical task research of rather wide interest.
If the guest substance is in a gaseous state and
slightly soluble in liquid water, some means to
enhance the mixing of the guest substance and
water is necessary to realize hydrate formation at a
rate high enough for industrial purposes.
Stirring is the most common method used to
improve mass transfer and heat transfer in the
methane hydration process: induction times was
shortened, formation rates accelerated, and storage
capacity was increased when stirring was applied
[10-11]. Stirring reactors are widely used in
laboratory settings to carry out fundamental
studies on formation and dissociation of gas
hydrates [10]. Anyway, as already established,
there are several problems with stirring in an
industrial-size process. Energy costs from stirring
increase as slurry thickens. Thickening slurry of a
stirred system may limit the hydrate mass in the
water to as low as 5% [12] at which time filtering
or clarification would be necessary. The separation
of hydrates from the slurry requires additional
work as does packing the hydrates in a storage
vessel. Furthermore, hydrate particles entrain
substantial amounts of interstitial water.
The bubble tower reactor is a traditional industrial
reactor type suitable for gas-liquid reactant system,
and it was tested for hydrate formation. So far,
results from use of this type of reactor are not
encouraging in terms of hydrate formation rate,
and hydrate shells formed are not easy to be
broken [10, 13].
Spraying reactor is another type of reactor for gasliquid reactant system. Application of water
spraying into a gas phase to form methane
hydrates has been studied by Mori and co-workers,
who designed and modified spraying reactors [1415]. We believe this technique is worthy of further
development for applications to scaled-up systems.
In this work we report and discuss the design and
development of a new scaled-up spraying reactor,
with inner volume of 25 l, with 6 nozzles. It was
designed to be suitable for practical engineering
use, for rapid and in continuo production of
hydrates. So far, it has been tested as batch reactor,
as will be shown in detail in the next section. The
objective of experimentation is to lower energetic
costs of hydrate formation, also through surfactant
promotion. We tried to decrease the overall time,
both induction time and hydrate-formation time: in
fact, time is an important factor to be taken into
account so that some authors use the expression
“time cost” [16]. At this regards, the use of
surfactants in hydrate-forming systems provides
an intriguing subject of research. Previous studies
reported that addition of surfactant to liquid water
causes a drastic promotion on the rate of formation
and occupancy of gas hydrates, and decreases the
induction time [9, 16-21]. In our work we used
anionic surfactant sodium dodecyl sulphate, SDS,
which seems one of the best commercially
available surfactant to be used for enhancement of
methane hydrate formation [17-19].
REACTOR DESIGN
A novel apparatus of internal volume of 25 l has
been designed and built to carry out studies on the
scale-up of gas hydrate formation. A schematic
diagram of it is shown in Figure 1.
Figure 1: Schematic diagram of the experimental
apparatus.
The reactor has been designed to maximize
interfacial area between reactants, thus minimizing
mass transfer barriers and thermal effects that
negatively affect conversion of reactants into
hydrate. Hydrates, in fact, are formed in the
reactor by injecting water through spray nozzles
from the top of the methane-filled reactor.
It consists of a high-pressure cylindrical AISI 304
stainless steel vessel with internal diameter of 200
mm, an internal length of 800 mm and a total
internal volume of 25 l. It has been designed for
pressure values up to 120 bar and provided with a
safety valve. A system was developed to remove
reaction heat and to ensure a rather constant
temperature inside the reaction vessel. The reactor
is provided with a screw conveyor. Thread depth
increases from flanges toward the center of the
vessel and is tangent to a finely-perforated metal
sheet. This system is designed to allow
conveyance of the solid hydrates to the unloading
flange, with a first separation of liquid water from
solid products. Two AISI 304 stainless steel
flanges are used to seal the reactor. One flange has
appropriate ports for access to the interior. The
four ports are used for inserting 2 temperature
sensors and for supplying methane and the
pressurized aqueous solution. The temperature
sensors are mineral insulated type T
thermocouples (accuracy class 1) and measure the
temperature inside the vessel in the lower part and
in the upper part, near spray nozzles (see scheme
1). The methane inlet line is equipped with a
pressure sensor, that is a digital piezo-resistive
manometer (accuracy class 0.5). Gas is supplied
directly by gas bottles through a pressure-reducing
valve, that provides adjustment of the pressure to
the methane injection line.
Water-supply system comprises the device for
water pressurization and cooling and the manifold
for its injection. The aqueous solution is flowed by
a circulation pump to a high-pressure vessel,
placed in a cylindrical water thermostatic bath.
The bath was equipped with an immersion type
cooler. The cooling medium is ethylene glycol –
water solution. The aqueous solution is pressurized
inside the vessel through injection of nitrogen and
sent to the manifold, equipped with pressure gauge
for the measure of water pressure before nozzles.
In the internal part of the manifold, water is
atomized by six hydraulic nozzles. Such devices
for water spraying allow control of dimensions of
water droplets and water flow. The temperatures
of the thermostatic bath was measured by a
thermocouple inside of the bath, and the
temperature of the injected aqueous solution was
measured by another thermocouple placed on the
inlet pipe.
Voltage signals from pressure transducers and
temperature sensors are collected by a software for
data acquisition on a personal computer.
The installation of the system is shown in Figure
2. The reactor is in the foreground. Seen in the
background there are gas bottles, the pressurized
water vessel, and liquid nitrogen cylinder.
Nitrogen (99.999 % purity) and methane (99%
purity) were supplied by Air Liquide Italia
Service. Distilled water was used to prepare
solutions.
and temperature vs time; data reported in this
figure are for experimental run 2 in Table 1. The
pressure is the total pressure inside the reactor, and
the temperature values are three, because two
values are monitored inside the reactor in two
different positions (see Figure 1) and one value is
relating to the temperature of sprayed water.
Figure 2: The methane hydrate formation reactor
(in the foreground), gas bottles, water vessel and
nitrogen cylinder (in the background).
APPLICATION OF THE REACTOR FOR
METHANE HYDRATE FORMATION WITH
SURFACTANT PROMOTION
We used the reactor described to produce methane
hydrates, but it can be used for other types of
hydrates (natural gas for instance). It was designed
to produce hydrates in a rapid manner, with
hydrate formation times below 1 hour. The
experimental procedure includes the preparation of
the system and the carrying out of the experiment
itself. Hydrate formation is carried out through the
injection of atomized aqueous solution into the
methane-filled reactor.
In the preparation, the reactor is filled with
methane from gas bottles until the internal
pressure equals the experimental pressure and then
cooled by liquid nitrogen. The temperature is
controlled in order to achieve relatively uniform
values inside the reactor. Water (aqueous
surfactant solution) is injected into the
pressurization vessel with a circulation pump and
cooled in the thermostatic bath. Then, it is
pressurized with nitrogen from gas bottles, until
the correct value of differential pressure on
nozzles is reached.
When the experimental conditions are reached,
aqueous solution is flowed through the nozzles.
Since the flow rate of aqueous solution through the
nozzles depends on the differential pressure on
nozzles themselves, the water spraying is
continued for several minutes until the established
total amount of aqueous solution is injected.
During the experiment pressure and temperature
data are collected every 5 seconds. Figure 3
represents typical profiles of values of pressure
Figure 3: The variation of values of pressure and
temperatures with elapsing time for methane
hydrate formation in SDS 300 ppm- experimental
run 2 in Table 1. Water spraying starts at t = 1
min.
As regards the temperature values, our efforts in
the reactor design were devoted to obtain an high
level of temperature control. Dealing with
temperature control is one of the important factors
deciding the performance of the hydrate formation
for at least two reasons: subcooling and heat
removal. In fact, under the experimental condition
of a pressure of 8 MPa, the equilibrium
temperature of methane hydrate is 287.0 K [1].
However subcooling is generally required for
hydrate formation so that it is often considered a
driving force for the process. Under certain limits,
formation rate increases with increasing
subcooling [21]. In our work the experiments were
carried out at initial temperature values of ca. 3-4
°C, with a relatively high subcooling. Furthermore,
during the hydrate formation the temperature
increases as a result of the exothermic process, and
it is of no help in hydrate formation, so that an
efficient system for reaction heat removal is
necessary. Moreover, in a scaled-up system there
is the further problem of temperature uniformity
inside the volume of the reactor. In Figure 3 we
can see how we were able to attain constant
experimental values of temperature and pressure in
the first minute (no water is present inside the
reactor in the first minute). After this time had
elapsed, water spraying started, and continued for
15 minutes. Temperature profiles reported in
Figure 3 show how our system for temperature
control was able to maintain a relatively constant
and uniform values of temperature in the reactor
during hydrate formation, with variations within
3°C.
As regards variation of pressure with time, one
would expect a pressure increase during water
spraying due to the decrease of available volume.
In Figure 3 it can be clearly noted that the pressure
increase tends to be linear at the beginning, in the
first 4-5 minutes. Afterwards, the pressure
variation tends to be smoother, and then it levels
off. This can be ascribed to gas consumption due
to hydrate formation. For clarity in Figure 4
pressure values during water spraying are reported
for experiments carried out with and without SDS
surfactant, i.e. runs 1 and 2 in Table 1.
Despite that, the apparatus is designed for
collecting gas hydrates through the activation of
the rotating screw. Therefore, at the end of the
experiment, gas was vented out from the methane
injection port through a three-way valve. Visual
observations was possible opening the flange, and
interesting images were taken.
Figure 5: Methane hydrate produced in our
reactor.
Figure 4: The variation of pressure with elapsing
time for methane hydrate formation with and
without SDS promotion - experimental runs 1 and
2 in Table 1. Water spraying starts at t = 0 min.
It can be clearly noted that in the absence of
surfactant the pressure increase is almost linear all
over the total water spraying time, indicating
almost no methane capture in the hydrate lattice.
From this comparison we may also deduce that in
presence of SDS in our experimental conditions
the induction time is almost negligible, and
anyway shorter than 4-5 minutes.
In this first set of experimental applications of the
reactor we chose to work in batch conditions, to
open the reactor and have a visual analysis of the
situation, and to take hydrate samples out from the
reactor for characterization in terms of gas content.
Figure 5 shows the picture of the hydrate formed
in run 2, after 15 minutes reaction time. It can be
clearly seen how there is no liquid water left. The
hydrate formed mainly on metal surfaces, such as
the lower part of the inner surface of reactor, the
outer surface of the screw conveyor, the upper part
of the finely-perforated metal sheet. This
phenomenon is consistent with those observed by
different authors, who reported that hydrates form
on the metal surfaces [17, 21-22]. Several samples
were taken directly out from the reactor since the
flange is open. Hydrate storage capacity was
determined putting hydrate samples inside a
custom built dissociation vessel, shown in Figure
6. It is a cylindrical AISI 304 stainless steel vessel
with a volume of 1.4 lt. It was designed and built
to carry out the dissociation of samples of gas
hydrate formed.
After sealing the vessel, the dissociation starts and
gas methane pressure and temperature after
dissociation are measured. To calculate number of
methane moles the following equation was used:
PV=ZnRT
(1)
where P is the methane partial pressure in the
vessel at the end of dissociation, V is the volume
of gas in the vessel, n is number of the gas
methane moles, T is the temperature in K at the
end of dissociation, R is the universal gas constant,
and Z is the compressibility factor, which can be
calculated using Benedict-Webb-Rubin equation
of state.
profiles in Figure 4, that without SDS the gas
content in the hydrate is very small.
Run
Initial
T, °C
SDS,
ppm
Water,
ml
Water
spraying
time, min
%CH4a
1
2
3
4
5
3
3
4
4
4
300
300
300
300
5070
5070
2500
6000
6000
15
15
15
25
25
<1 (1)
4 (6)
4 (8)
6 (8)
2 (4)
a
weight %; mean value among several samples; maximum value in
parentheses.
Table 1: Summary of the results of the
experimental application of the novel reactor for
methane hydrate production.
Figure 6: The dissociation vessel.
As the number of gas moles is calculated, hydrate
storage capacity, measured in % of weight, can be
determined, because sealed vessel had been
weighted right after closure. The maximum value
of methane storage capacity in methane hydrate of
structure-I is 13.4 in weight %.
In Table 1 we report the results of our first set of
experimental applications of the reactor for rapid
methane hydrate formation. As initial temperature
we indicate the mean value between values
measured in the upper and lower part of the reactor
before starting water spraying. In the formation of
gas hydrate by adding SDS, the decision of the
ideal concentration of SDS is primarily required.
Several investigations reported in the literature
indicate different optimal SDS concentrations, in
different conditions and reactors [9, 16-17].
Anyway a concentration in the range of 250-500
ppm seems the best choice: as a starting
concentration for a first set of experiments we
chose 300 ppm. We also carried out an experiment
in the absence of SDS for the sake of comparison.
First of all, it is clear, as already outlined from
In presence of SDS, in general, we have to say that
the amount of methane stored in the hydrates is not
high compared to literature values, which refer to
laboratory scale reactors. It was expected, because
the scale-up effects studied so far indicate that the
hydrate formation rate, the moles of gas consumed
per unit mass of water and time decreased rapidly
with increasing the mass of water loaded in the
reactor [10, 22]. Our result confirms this general
trend, but it is necessary to point out that our
reaction time is quite short, even compared to
laboratory scale experiments where times of
hundreds of minutes are used and our water
volumes are significantly greater than the amounts
used in the laboratory scale, that are few hundred
ml or less [9-21]. We also have to point out that
we never found liquid water left in our
experiments. Due to the fact that large scale
industrial applications can not face long reaction
times, we tried to improve the performance
changing conditions but without lengthening so
much the time. A decrease in the water amount
from ca 5 l to ca. 2.5 l did not change so much the
situation (runs 2-3). We tried to leave time to the
system for hydrate formation to occur after
finished water spraying, but almost nothing
happened after water spraying had finished (data
not shown in table 1). On the other hand, the
methane storage increased if the spraying time was
lengthened from 15 to 25 minutes, even with a
greater amount of water (runs 2-4). A modification
in the reactor, with momentary elimination of the
screw conveyor, gave a worse performance, and it
can be related to a decrease in the metallic surface
available (runs 4-5).
In conclusion, a novel reactor, with a volume of 25
l, for the production of methane hydrates was
designed, constructed and tested for its application
to rapid hydrate formation, with the word rapid
meaning a reaction time of some tens minutes,
instead of the hundreds minutes used in the
literature also in the laboratory scale. Results of a
first set of applications indicate that we could
obtain an high level of temperature control in the
reactor during hydrate formation, and we could
obtain methane hydrates, without water left, in
some tens minutes only with SDS surfactant
promotion. The amount of water loaded was up to
6 l. Changing some experimental conditions we
improved the gas content in the hydrate, that is not
generally high: the best value we have obtained is
ca. half the maximum. On the basis of these
results, our application of the reactor is continuing.
Other variations in the conditions are planned, and
slight variations in the reactor design, via
interchangeable pieces, are under development in
order to obtain an higher level of gas content in the
rapidly formed methane hydrates.
REFERENCES
[1] Sloan, E. D. Clathrate hydrates of natural
gases. New York: Marcel Dekker, 1998.
[2] Handa Y. Calorimetric determination of the
composition, enthalpies of dissociation and heat
capacities in the range of 80-270 K for clathrate
hydrates of Xenon and Kripton. J. Chem. Therm.
1986;18:891-903.
[3] Gudmundsson JS, Khodakar AA, Parlaktuna
M. Storage of natural gas as frozen hydrate. In:
Proceedings of the sixty-seventh annual technical
conference and exhibition of SPE 1990, p. 699707.
[4] Gudmundsson JS, Brrehaug A. Frozen hydrate
for transport of natural gas. In: Proceedings of the
2nd International conference on natural gas
hydrates, Toulouse, 1996, p. 415-422.
[5] Shirota H, Aya I, Namie J. Measurement of
methane hydrate dissociation for application to
natural gas storage and transportation. In:
Proceedings of the fourth International conference
on natural gas hydrates, Yokohama, 2002, p. 972977.
[6] Takaoki T, Iwasaki T, Katoh Y, Arai T,
Horiguchi K. Use of hydrate pellets for
transportation of natural gas. In : Proceedings of
the fourth International conference on natural gas
hydrates, Yokohama, 2002, p. 982-986.
[7] Kanda H. Economic study on natural gas
transportation with natural gas hydrate (NGH)
pellets. In: Proceeding of the 23rd world gas
conference, Amsterdam 2006.
[8] Rossi F, Cotana F, Castellani B, Filipponi M.
Nuovo reattore sperimentale per la produzione di
gas idrati.
Proceedings of the X National
Congress CIRIAF, Perugia 2010.
[9] Okutani K, Kuwabara Y, Mori YH. Surfactant
effects on hydrate formation in an unstirred
gas/liquid system: An experimental study using
methane and sodium alkyl sulfates. Chemical
Engineering Science 2008;63:183-194.
[10] Lang X, Fan S, Wang Y. Intensification of
methane and hydrogen storage in clathrate
hydrate and future prospect. Journal of Natural
Gas Chemistry 2010;19:203-209.
[11] Hao W, Wang J, Fan S, Wenbin, H. Study on
methane hydration process in a semi-continous
stirred tank reactor. Energy conversion
Management 2007;48:954-960.
[12] Vysniauskas A, Bishnoi PR. A kinetic study
of methane hydrate formation. Chemical
Engineering Science 1983;38:1061-1072.
[13]Luo YT, Zhu JH, Fan SS, Chen GJ. Study on
the kinetic of hydrate formation in a bubble
column.
Chemical
Engineering
Science
2007;62:1000-1009.
[14] Ohmura R, Kashiwazaki S, Shiota S, Tsuji H,
Mori YH. Strucure-I and structure-H hydrate
formation using water spraying. Energy Fuel
2002;16:1141-1147.
[15] Tsuji H, Ohmura R, Mori YH. Forming
structure-H hydrates using water spraying in
methane gas: effect of chemical species of largemolecule guest substances. Energy Fuel
2004;18:418-424.
[16] Sun ZG, Wang R, Ma R, Guo K, Fan S.
Natural gas storage in hydrates with the presence
of
promoters.
Energy
Conversion
and
Management 2003;44:2733-2742.
[17] Zhong Y, Rogers RE. Surfactant effects on
gas hydrate formation. Chemical Engineering
Science 2000;55:4175-4187.
[18] Daimaru T, Yamasaki A, Yanagisawa Y.
Effect of surfactant carbon chain length on
hydrate formation kinetics. J. Petroleum Sci and
Eng. 2007;56: 89-96.
[19] Ganji H, Manteghian M, Sadaghiani zadeh K,
Omidkhah MR, Rahimi Mofrad H. Effect of
different surfactants on methane hydrate formation
rate, stability and storage capacity. Fuel
2007;86:434-441.
[20]Di Profio P, Arca S, Germani R, Savelli G.
Surfactant promoting effect on clathrate hydrate
formation: are micelles really involved Chemical
Engineering Science 2005;60:4141-4145.
[21]Lee J, Shin C, Lee Y. Experimental
investigation to improve the storage potentials of
gas hydrate under the unstirring conditions.
Energy Fuel 2010;24:1129-1134.
[22] Pang WX, Chen GJ, Dandekar A, Sun CY,
Zhang CL. Experimental study on the scale-up
effect of gas storage in the form of hydrate in a
quiescient reactor. Chemical Engineering Science
2007;62:2198-2208.